ISSUES AND CONTROVERSIES

Dose Optimization to Minimize Radiation Risk for Children Undergoing CT and Nuclear Medicine Imaging Is Misguided and Detrimental

Jeffry A. Siegel1, Bill Sacks2, Charles W. Pennington3, and James S. Welsh4

1Nuclear Physics Enterprises, Marlton, New Jersey; 2U.S. Food and Drug Administration (retired), Green Valley, Arizona; 3NAC International (retired), Norcross, Georgia; and 4Department of Radiation Oncology, Stritch School of Medicine, Loyola University, Maywood, Illinois

for “Radiation Dose Optimization in Pediatric Nuclear Medicine” A debate exists within the medical community on whether the linear advertises 13 presentations on radiation risk in children, absorbed no-threshold model of ionizing radiation exposure accurately dose estimation, dose reduction, and radiopharmaceutical guidelines. predicts the subsequent incidence of radiogenic cancer. In this This seminar will focus on communicating a risk that is, in fact, article, we evaluate evidence refuting the linear no-threshold model fictitious; on emphasizing a dose reduction that is, consequently, and corollary efforts to reduce radiation exposure from CT and needless; on achieving a resulting dose estimation enhancement that nuclear medicine imaging in accord with the as-low-as-reasonably- is superfluous; and on improving activity-administration guidelines. achievable principle, particularly for children. Further, we review studies demonstrating that children are not, in fact, more radiosen- The underlying intent of lowering future cancer risk, although desir- sitive than adults in the radiologic imaging dose range, rendering able, goes astray, as the premise is based on the erroneous LNT and dose reduction for children unjustifiable and counterproductive. the resulting as-low-as-reasonably-achievable principle (ALARA). Efforts to minimize nonexistent risks are futile and a major source of The seminar further aims at procedure standardization in pediatric persistent radiophobia. Radiophobia is detrimental to patients and nuclear medicine, with which we have no quarrel. Although more parents, induces stress, and leads to suboptimal image quality and accurate dosimetry can improve guidelines for radiopharmaceutical avoidance of imaging, thus increasing misdiagnoses and conse- activity administration and is a reasonable pathway to optimization quent harm while offering no compensating benefits. of image quality, it is wholly misguided for radiation protection Key Words: radiological imaging; linear no-threshold; ALARA; dose purposes. optimization; radiophobia; children LNT and its corollary, ALARA—which are based on a link J Nucl Med 2017; 58:865–868 between absorbed radiation (no matter how low the dose or dose DOI: 10.2967/jnumed.117.195263 rate) and future cancer probability—are defended despite their proponents’ disarming admission that because of the low-dose- radiation signal-to-noise problem, there is no evidence, nor can there be any, favoring LNT for low doses. Proponents often defend their reliance on LNT on the grounds that even if there is a dose There is a great divide over exposure to low-dose radiation. threshold below which there is no harm, LNT’s consequent over- Ionizing radiation is thought by many to cause cancer, based on estimation of harm errs on the side of caution. Seemingly sound, uncritical acceptance of the linear no-threshold model (LNT) de- this application of the precautionary principle ignores empiric spite absence of valid evidence for doses below 100–200 mGy (1). low-dose-radiation research while offering no refutation, as well Moreover, the LNT extrapolation from high- to low-dose radiation as denying responsibility for the radiophobia, with its attendant has proven empirically false and decidedly harmful to millions of harms, generated by LNT and the concomitant result that patients people with respect both to medicine and to government policies are deprived of the probable cancer-inhibiting effects of low-dose regarding nuclear power plants (1–4). Herein, we argue that low- radiation (1,4). dose and low–dose-rate radiation does not cause—but in most The precautionary principle is useful only if action to control people helps prevent—cancer, as shown by sound, empiric, exper- the feared agent has no, or less harmful, side effects. However, for imental and observational studies. We also briefly summarize the radiologic imaging, the negative consequences of radiophobia- research showing the falsehood of LNT and its detrimental con- based dose reduction exceed the putative harms of radiation. sequences. These consequences include misdiagnosis due to both imaging This article was prompted by an all-day categoric seminar that will avoidance and nondiagnostic image quality, as well as use of take place this month at the SNMMI annual meeting. The program alternative imaging procedures, such as longer-duration MRI studies, that have a higher sedation risk for children, as we have reported (1). The harmful consequence of misdiagnosis—the need Received Apr. 24, 2017; accepted Apr. 24, 2017. For correspondence contact: Jeffry A. Siegel, Nuclear Physics Enterprises, to resort to exploratory surgical alternatives—is greater than the 4 Wedgewood Dr., Marlton, NJ 08053. LNT-based fictitious prediction of harm and more immediate than E-mail: [email protected] Published online May 11, 2017. any future cancer risk that is putatively avoided by ALARA dose- COPYRIGHT © 2017 by the Society of Nuclear Medicine and Molecular Imaging. reduction strategies (5).

MISGUIDED PEDIATRIC DOSE OPTIMIZATION • Siegel et al. 865 A worldwide movement has arisen in response to the scientif- stimulate the immune system to reduce cancer rates (9). Fahey ically unsupportable call to decrease dose. We concentrate here on et al., nevertheless, claim it is prudent to strive for optimal ad- the pediatric CT and nuclear medicine dose “optimization” efforts ministered activity in children by using a patient-size–related generated by the Image Gently Alliance for children, a more so- dose, defining optimal as the lowest dose that still provides the cially and medically vulnerable population. The efforts aim not diagnostic information necessary for proper care (10). Although only at improving diagnostic quality, a laudable objective, but also an insufficient dose has resulted in nondiagnostic studies, this at lowering the fictitious risk of future cancer. Recognizing the trade-off ignoring the body’s proven defensive responses is antagonism between these contradictory objectives, Dr. Donald thought to be necessary. Frush, chair of the Image Gently Alliance, states that the Alliance As with the study of Preston et al. (8), much of the current seeks to provide balance between them, yet he disclaims any re- thinking about the risks of ionizing radiation is based on certain sponsibility of the Alliance for the media-driven radiophobia (6). interpretations of the Hiroshima/Nagasaki life-span study. As we Reasonable as this aim might be if, in fact, there were a risk of have reported (1), these data, properly interpreted, invalidate LNT. cancer from CT or nuclear medicine imaging, balance between Supportively, the French Academy of Sciences (11) reported that diagnostic benefit and cancer risk can no longer claim innocence these data provide evidence for protective, adaptive responses and in the absence of such risk. Balance between truth and falsehood is no valid evidence for harm below 100 mGy. Despite this contrary no balance at all. evidence, report VII from the National Academy of Sciences’ Committee on the Biological Effects of Ionizing Radiation (BEIR CHILDREN ARE NOT MORE RADIOSENSITIVE THAN ADULTS committee) (12) still endorsed LNT for solid cancers and, further, attributed a risk 2–3 times higher in children. Unaware perhaps of The focus on children is founded on their social vulnerability and the BEIR committee’s distortion of its own cited data, most radi- on a false biologic assertion that children are more radiosensitive ologists and nuclear physicians look to this widely regarded voice than adults. This assertion is based both on an indisputable fact and of authority. Indeed, we have found and revealed BEIR VII’s on a false assumption: the fact that more rapid cell division leads to egregious misquoting of one of its chosen sources that actually greater mutation opportunities, and the assumption that increased came to the opposite conclusion, in line with our contention (4). mutations entail increased cancer risk. But, like LNT itself, the as- sumption ignores evidence-supported adaptive responses that either VOICES OF “AUTHORITY” DO MORE HARM THAN GOOD repair mutations through enhanced repair enzymes or remove the unrepaired cells by apoptosis or, most importantly, the immune Unfortunately, many imaging professionals do not read scien- system (1,3,7). In the absence of these evolution-provided naturally tific literature critically. For their understanding of LNT, ALARA, selected defenses, the far more frequent (by about a million-to-one) and dose optimization, they rely on “expert” opinion, as offered in endogenous mutations in almost every cell—due to reactive oxygen reports by the BEIR committee, the International Commission on species from normal metabolism—would produce cancer minute Radiological Protection, the National Council on Radiation Pro- by minute. Mutations or epigenetic changes are necessary, but not tection and Measurements, and others that mutually reinforce each sufficient, to produce clinical cancer. other. Further, many investigators discount the work of those with Most importantly, Preston et al. studied Japanese atomic bomb whom they disagree rather than even attempting to refute it, in- survivors who were younger than 6 y at the time of the bombing (8). stead appealing to citations from these authoritative organizations The results indicated no significant difference in adult-onset solid- as though they were established fact rather than the mere opinion cancer incidence between the control group and the group that re- that closer examination proves them to be. ceived acute exposure of up to 200 mSv as children. CT and nuclear As we have reported (1,4), there are two large epidemiologic medicine imaging apply doses that are at least an order of magni- studies that have been frequently cited as evidence to support the tude below that figure. Thus, there is epidemiologic evidence that claim that children are at an increased risk of cancer due to CT. children are not more radiosensitive to such imaging. But according to the United Nations Scientific Committee on the In addition, relevant evidence from a study on individuals Effects of Atomic Radiation, and even Dr. John Boice (13), pres- undergoing CT examination indicates that at low radiation expo- ident of the National Council on Radiation Protection and Mea- sures, initial radiation-induced damage is generally repaired or surements, the results of these studies are likely untrue. Inaccurate eliminated in a matter of hours by the body’s adaptive responses. dosimetry and implausible risk estimates are apparent; association Löbrich et al. (7), for example, showed that, within minutes, expo- is conflated with causation; and medical indication rather than sure to radiation from CT scans induces an increase in DNA double- radiation is likely the source of the apparent association (i.e., re- strand breaks compared with pre-CT levels. But these double-strand verse causation). Further, again as we reported and as underscored breaks are subsequently repaired. In fact, in all but one patient, by Dr. Boice, two more recent epidemiologic studies, in France the double-strand breaks were repaired to less than the initial and Germany in 2015, found no significant excess cancer risk in (pre-CT) background level at some time between 5 and 24 h, children from CT scans, “...once adjustment was made for con- suggesting that the low-dose CT exposure induced repair of the ditions that prompted the scan, family history, or other predispos- preexisting and ongoing endogenous DNA damage as well as ing factors known to be associated with increased cancer risk” repair of the radiogenic damage. This is evidence of a beneficial (13). Thus, confounding by indication (reverse causation) must (hormetic) effect of low-dose ionizing radiation and argues be ruled out completely, or observed excess cancer risk may be against radiogenic causation of either solid cancer or leukemia falsely, and facilely, attributed to CT exposure. in children or adults. Dr. David Brenner et al. (14) published an article in 2001 that Furthermore, in older children the immune system, once was widely promoted by the media, accelerating the prevalence established, is more vigorous than that in adults; its efficiency and intensity of radiophobia. This article has been cited in the declines with age, and low-dose radiation has been shown to scientific literature over a thousand times since. On the basis of

866 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 58 • No. 6 • June 2017 an LNT calculation involving unsupported parameters with signif- effective dose, therefore, is well within the hormetic zone; that icant uncertainties, the authors projected that approximately 500 is, below the threshold for harm. children under the age of 15 y would eventually die of CT-caused The original aim of the radiopharmaceutical guidelines was to cancer. Some years later, in the New England Journal of Medicine, reduce the variability of administered activities. This aim is praise- the same Brenner, with Hall, predicted that up to 2% of all cancers worthy for better standardization of image quality but irrelevant for in the United States “could” be due to CT scans (15). Importantly, radiation protection purposes. The need for better accuracy in dose these erroneous and irresponsibly fear-amplifying predictions estimation to further minimize a nonexistent risk based solely on the were based on CT doses that were completely unknown for individ- erroneous LNT is entirely misdirected and, as a serious unintended ual patients and were based on risk estimates from the Hiroshima/ consequence, serves only to amplify long-standing public and pro- Nagasaki life-span study cohort. These risk estimates are not ap- fessional radiophobia. plicable to children undergoing CT examinations in the United States, for one thing because it is most unlikely that the radiation will have played a causal role in children who do develop cancer CONCLUSION after a CT scan. Further, whatever the cancer rates in bomb sur- Many critics argue for termination of the fictitious LNT, along vivors, they were also subjected to the confounding factors of with its epiphenomenal ALARA and the Image Gently Alliance, severe trauma and deprivation, which are known to interfere with but must contend with a well-ensconced paradigm—one backed immune system function and hence enhance susceptibility to all by worldwide advisory organizations and regulatory agencies sorts of ailments. whose overlapping members have invested reputations and labor. Brenner’s group subsequently reported wholly unjustified con- Critics must fight not only for predominance of empiric evidence clusions again promoting widespread radiophobia, this time with but also against concepts such as the widely accepted yet mis- respect to double-strand breaks in only 3 pediatric patients after applied justification and optimization principles of the Interna- undergoing CT examinations (16). Even acknowledging the find- tional Commission on Radiological Protection (1). Eliminating ing of Löbrich et al. (7)thatrepairorremovalofinitialdam- any and all diagnostic medical procedures that are not clinically age was expected within several hours of exposure, this group warranted is important. But attempting to lessen fictitious risk by obtained data only at 1 h after the CT scans, which, as expected, lowering dose in studies that are clinically warranted is a mis- demonstrated increased double-strand breaks compared with pre- application of the principles. It is protection from radiophobia, CT levels. Yet the group concluded that its findings supported rather than from low-dose radiation, that will benefit both chil- LNT at very low doses in young children and that “even very low dren and adults. ionizing radiation exposure relevant to diagnostic CT exposure The Image Gently Alliance and myriad international affiliated can leave a mark in the somatic DNA,” significantly and irre- groups harm patients and parents by promoting radiophobia while sponsibly omitting the qualifier “for perhaps a few hours.” offering fictitious protection against nonexistent risks. We do not Astonishingly, the group recommended that, based on this ad- suggest a cavalier approach by discounting dose optimization, but mittedly incomplete evidence, “When possible, CT exams should only in service of diagnostic quality. Public misperceptions should be limited or avoided by possibly applying non-ionizing radia- not be disregarded or ignored, but neither should they be endorsed; tion exposure techniques such as US or MRI.” Such alternatives rather, they must be corrected through accurate information. The are sometimes medically warranted, but their use should be public’s trust in medical practitioners can be attained only if we based only on their being the most appropriate imaging modality, convince the public and physicians alike that there is no harm in not on their producing a lower radiation exposure. radiologic imaging. Accurate information about low-dose radiation is the only way to THE SEARCH FOR ACCURACY IN THE ABSENCE OF VALIDITY undo fear, as decades of failed alternative approaches and conces- sions have shown. Although it may seem reasonable to attempt to The dose-optimization movement in nuclear medicine has assuage public fears by accommodating misperceptions through prompted intense de novo efforts to develop numerous complex lowering dose despite the absence of harm, such measures only anatomic phantoms and better pharmacokinetic data and models reinforce the misperceptions. Radiophobia is detrimental to patients to more accurately estimate absorbed doses for diagnostic proce- and parents, induces stress, and leads to suboptimal image quality or dures (as opposed to therapeutic procedures) (17). Furthermore, avoidance of imaging, both of which produce misdiagnoses. Mis- the current consensus guidelines for administered radiopharma- perception can be overcome only by rejection of the LNT fiction and ceutical activity are considered inadequate and in need of updat- its corollary principle, ALARA, and by termination of the Image ing. This more precise and labor-intensive endeavor is deemed Gently Alliance. necessary to improve the practice of pediatric nuclear medicine imaging by providing excellent image quality at the lowest radi- ation dose possible, but the obsession over lowering radiation dose DISCLOSURE is a futile and laborious attempt to minimize what is, in fact, a nonexistent risk. No potential conflict of interest relevant to this article was As one example of the relative superfluity of such precision, reported. much effort has already been expended on assessing pediatric use of 99mTc-dimercaptosuccinic acid, for which the highest effective dose per imaging examination is only about a third of the average REFERENCES annual natural background radiation dose—3.1 mSv—received 1. Siegel JA, Pennington CW, Sacks B. Subjecting radiologic imaging to the linear per person in the United States, with the average background dose no-threshold hypothesis: a non sequitur of non-trivial proportion. J Nucl Med. elsewhere in the world ranging to up to 80 times greater. This 2017;58:1–6.

MISGUIDED PEDIATRIC DOSE OPTIMIZATION • Siegel et al. 867 2. Siegel JA, McCollough CH, Orton CG. Point/counterpoint: advocating for use of 10. Fahey FH, Ziniel SI, Manion D, Treves ST. Effects of Image Gently and the the ALARA principle in the context of medical imaging fails to recognize that North American guidelines: administered activities in children at 13 North the risk is hypothetical and so serves to reinforce patients’ fears of radiation. Med American pediatric hospitals. J Nucl Med. 2015;56:962–967. Phys. 2017;44:3–6. 11. Aurengo A, Averbeck D, Bonnin A, et al. Dose Effect Relationships and 3. Siegel JA, Welsh JS. Does imaging technology cause cancer? Debunking the Estimation of the Carcinogenic Effects of Low Doses of Ionizing Radiation. linear no-threshold model of radiation carcinogenesis. Technol Cancer Res Treat. Paris, France: Académie des Sciences–Académie nationale de Médecine; 2016;15:249–256. March 2005. 4. Sacks B, Meyerson G, Siegel JA. Epidemiology without biology: false para- 12. National Research Council of the National Academies. Health Risks from Expo- digms, unfounded assumptions, and specious statistics in radiation science. Biol sure to Low Levels of Ionizing Radiation: BEIR VII Phase 2. Washington, DC: Theory. 2016;11:69–101. The National Academies Press; 2006. 5. Siegel JA, Stabin MG. Medical imaging: the challenges of radiation risk assess- 13. Boice JD Jr. Radiation epidemiology and recent paediatric computed tomogra- ment. J Nucl Med. 2014;55(12):16N–17N. Ann ICRP. 6. Frush DP, Goske MJ. Image Gently: toward optimizing the practice of pediatric phy studies. 2015;44(1, suppl):236–248. CT through resources and dialogue. Pediatr Radiol. 2015;45:471–475. 14. Brenner D, Elliston CD, Hall EJ, Berdon WE. Estimated risks of radiation in- 7. Löbrich M, Rief N, Kühne M, et al. In vivo formation and repair of DNA double- duced fatal cancer from pediatric CT. AJR. 2001;176:289–296. strand breaks after computed tomography examinations. Proc Natl Acad Sci 15. Brenner DJ, Hall EJ. Computed tomography: an increasing source of radiation USA. 2005;102:8984–8989. exposure. N Engl J Med. 2007;357:2277–2284. 8. Preston DL, Cullings H, Suyama A, et al. Solid cancer incidence in atomic bomb 16. Halm BM, Franke AA, Lai JF, et al. g-H2AX foci are increased in lymphocytes survivors exposed in utero or as young children. J Natl Cancer Inst. 2008;100:428– in vivo in young children 1 h after very low-dose x-irradiation: a pilot study. 436. Pediatr Radiol. 2014;44:1310–1317. 9. Liu SZ. Cancer control related to stimulation of immunity by low-dose radiation. 17. Fahey FH, Goodkind AB, Plyku D, et al. Dose estimation in pediatric nuclear Dose Response. 2006;5:39–47. medicine. Semin Nucl Med. 2017;47:118–125.

868 THE JOURNAL OF NUCLEAR MEDICINE • Vol. 58 • No. 6 • June 2017